HANG drcachesim online across setarch execve

This surprised me b/c I thought we had multi-process tests including execve, so there is likely something particular about this specific execve?

I just wanted to disable ASLR for the app and it hangs:

$ bin64/drrun -t drcachesim -- setarch `uname -m` -R suite/tests/bin/simple_app
<Starting application /usr/bin/setarch (6420)>
<Initial options = -no_dynamic_options -client_lib '/home/bruening/dr/git/build_x64_dbg_tests/bin64/../clients/lib64/debug/libdrmemtrace.so;0;' -code_api -stack_size 56K -signal_stack_size 32K -max_elide_jmp 0 -max_elide_call 0 -early_inject -emulate_brk -no_inline_ignored_syscalls -native_exec_default_list '' -no_native_exec_managed_code -no_indcall2direct >
<Paste into GDB to debug DynamoRIO clients:
set confirm off
add-symbol-file ...>
<curiosity: rex.w on OPSZ_6_irex10_short4!>
<get_memory_info mismatch! (can happen if os combines entries in /proc/pid/maps)
	os says: 0x00007f8597254000-0x00007f8597277000 prot=0x00000001
	cache says: 0x00007f8597254000-0x00007f859730d000 prot=0x00000001
>
<(1+x) Handling our fault in a TRY at 0x00007f8597a8d93b>
<-- execve suite/tests/bin/simple_app -->
<Starting application /home/bruening/dr/git/build_x64_dbg_tests/suite/tests/bin/simple_app (6420)>
<Initial options = -client_lib '/home/bruening/dr/git/build_x64_dbg_tests/bin64/../clients/lib64/debug/libdrmemtrace.so;0;' -code_api -stack_size 56K -signal_stack_size 32K -max_elide_jmp 0 -max_elide_call 0 -early_inject -emulate_brk -no_inline_ignored_syscalls -native_exec_default_list '' -no_native_exec_managed_code -no_indcall2direct >
<Paste into GDB to debug DynamoRIO clients:
set confirm off
add-symbol-file ..>
<curiosity: rex.w on OPSZ_6_irex10_short4!>
<get_memory_info mismatch! (can happen if os combines entries in /proc/pid/maps)
	os says: 0x00007ffff75df000-0x00007ffff7602000 prot=0x00000001
	cache says: 0x00007ffff75df000-0x00007ffff7698000 prot=0x00000001
>
<(1+x) Handling our fault in a TRY at 0x00007ffff7e1c93b>

It just hangs there trying to read the pipe. So sthg with the pipe is not handled properly across that execve it seems. Not sure how other tests w/ execve differ.

In any case the workaround here is to run drrun under the setarch.